Just sharing what I came up with to solve the problem of ducting fresh air to a larger-than-stock-sized SMIC. It's not intended to be a detailed "how to build" thread, but rather provide some insights to others who many need to tackle a project like this. My goal was to cram the largest IC core I could in a SMIC configuration, so I could restore an A/C system (without resorting to going V-mount), and keep the normal sized battery in the stock location. This limited my turn-key options to the smaller Pettit kit (that doesn't require relocating the battery) or the similar M2 "Medium" IC kit. Had no luck finding either complete kit (w/duct) used, so I decided on going the DIY route.

I decided on one of these XSPower IC's, which I determined was the largest size core I can fit in the space available for a SMIC, and looks to be a copy of the core used in the M2 medium kit. More info on the IC core here: FD3S_SMIC_KIT, RX7 3RD GEN 93-97 SMIC JDM JUN STYLE, Mazda Single Intercoolers, XS Power Inc.

Since I didn't need the whole kit (piping would be useless for my single turbo with DBW throttle anyway), I searched & found a retailer on Ebay that sold me the exact same XSPower core by itself for about $170 shipped. So now it was on to figure out how to fabricate a custom duct to get air to it, as I had zero stock IC ducting pieces to even start with.

I got a good deal purchasing a used Pettit duct from a forum member, knowing upfront that it wouldn't fit my larger IC core, but I figured I'd at least be able to use the "snout" portion of it to pull air from the opening between radiator & bumper support, and graft that onto a sheet metal fabricated section for the back side to cover the IC surface.

Picture of the Pettit duct on my IC core as-is:



And now a picture of my "version 1" IC duct that is made of sheet AL and the grafted snout section of the Pettit duct:




While this version worked, it was not up to my standards, and didn't fit quite right - there are gaps between the IC face and the duct that you can't see in the picture here. So when it was time to hibernate the FD for winter projects and keep it off the salted roads, design & construction on the Version 2 duct started.

Here's a picture of the new duct, after the sheet AL fabrication and test fitting was done. Fits perfectly and covers the entire face of the IC, and maximizes air inlet path from what's available.



To get to this point, here's the gist of my build process:
1. Build rectangular "frame" of 1/2"x1/2" angle AL and flat 1/2" AL bar stock, 1/16" thick, to the dimensions of the IC core face. Used the bar stock for the bottom edge because the angle stock would have pushed the lower edge out too far, and cause interference issues with the OEM fans (lesson learned from Ver.1!). That provides a datum reference for the IC side, and a solid place to rivet the sheet AL metal skin to.
2. Taking measurements of the available hole between radiator & bumper opening, build a box section out of AL sheet metal - turns out a USPS Priority mail small box is an almost perfect template for this piece. I ended up making it slightly larger as there was extra room there to maximize cold air intake. The 4 corners of this piece as installed and poking up from the bumper support forms the datum points for the where the corresponding corners of the IC "frame" piece connect to.
3. From here on out it was an iterative process. Started with the bottom panel, which began as a large rectangle, strategically cut & folded along lines to get the corresponding points to connect. This picture was the bottom panel, with lower side panel sections taking shape:



4. The side & top panels were done in similar iterations. Working with a 3"x3" piece of AL sheet I picked up a local hardware store, I had plenty of stock left over when done, even with all the wasted metal and iterative cutting & trimming.

Ideally, I would have preferred to build this duct out of fiberglass, as it is a better heat insulator than aluminum, but building up a throw away mold for a fiberglass part is a major project in itself, and can get messy & expensive really quick. So now I figured I needed to find a coating to apply to the aluminum exterior that would provide some thermal insulation and look OEM-ish cosmetically. For that purpose, I decided on a Rustoleum truck bed liner coating - it's a water based brush-on coating, has a rough/sandy texture and can be had at Walmart for about $18/quart. I applied 3 coats of the stuff, and this is the end result:




As for the thermal insulation the stuff provides, I did a quick test with a heat gun to estimate that. First I took a bare piece of the same AL sheet metal stock and pointed my heat gun (at low setting) to one side of it, started a timer and then touched the opposite side of the metal - the time it took for it to be too hot to touch was my baseline for the bare AL. That time was approximately 10 seconds. I then repeated this test with the finished duct, pointing the gun at that large outside lower/bottom surface panel while keeping the gun at the same distance from the surface as the prior bare metal test. With the coating applied, it took approximately 40 seconds before the opposite side became too hot to touch, so I would conclude the coating is providing at least some measurable thermal insulation. All in, the Version 2 duct cost me about $50 in materials (<$30 in AL stock, and $18 for the coating), and enough hours to keep me busy and out of trouble.

When I'm back on the road again with the FD, it will be interesting to see where the IAT's look like in the logs.

 

nice job! I may tackle something similar but for oil my coolers, I have CF ducts for 19row coolers that I need to adapt to 25row coolers.

what's this here?

 

Nice work Pete... she getting some great attention.

Steve

 

Thanks, oil cooler ducting would be pretty easy to do with sheet AL - assuming the coolers are in the OEM locations. ^That would be a relay box I picked up on Amazon to house my 4 fan relays. When I got the car it already had a Link G4+ Fury ECU that the PO wired in; he did a pretty decent job on the wiring harness, but he didn't include provisions for some of the features the Link could support that I wanted (e.g., OMP, supporting 3 fan speeds + A/C, etc.), so I've since done a complete rewire job. The PO had the fans wired to come on individually at full speed with 2 relays, using 2 AUX outputs from the Link. Fortunately, since the OEM fan wiring was all there and unmolested, it was easy enough to reverse engineer the OEM fan system to operate as designed, using the Link to control them. So now both fans run together at one of the 3 possible speeds (low/med/high) based on coolant temp and whether the A/C is on/off.

I may end up relocating that relay box, because it's currently in the way of another cold air duct I want to add to direct outside air towards the air filter, which currently is just a shorty cone on the turbo.

 

Thanks Steve, it certainly kept me plenty busy during the COVID slow times!

 

Looks awesome, Pete!

I think the only reason to do fiberglass is it's marginally lighter and you can more easily get a more complex shape. If you are a race car shop and have all the stuff handy to set up and work with fiberglass that's a no brainer. Your method probably would take longer than a pro and fiberglass but you didn't need a bunch of fiberglass resin, mat, making a buck, all that crap.

I seriously doubt heat on the duct will be any sort of issue. You have constant fresh air coming in through the duct and into the core - the IC core itself will do all the work and the air will remove the heat. Just as long as the duct is designed well so it will get the most air to the IC you're in good shape.

Sounds like a great winter project!

Dale